Stroke remains a major public health problem with a disproportionate impact on African Americans and Hispanics. Our knowledge of risk factors has improved, while only limited data are available regarding genetic influences. Given that stroke is a complex disease, evaluation of risk factor phenotypes with greater heritability may reduce complexity, facilitate gene discovery, and refine stroke prevention strategies. The goal of this application is to evaluate the heritability of new vascular risk factors (homocysteine, carotid intima-media thickness (IMT), carotid distensibility, brachial endothelial reactivity, and left ventricular mass), among the families of high-risk Caribbean Hispanics from the Northern Manhattan Stroke Study cohort (NOMASS). We propose to systematically measure these quantitative phenotypes, assemble a DNA family study databank, perform a genome scan, and begin to identify quantitative trait loci through preliminary linkage analysis. This will be the first family study focused on the rapidly growing Caribbean Hispanic race-ethnic group. A family study is proposed among 100 high-risk Caribbean Hispanic families (1500 subjects) selected and enrolled over four years. The enrollment and baseline data collection will be performed as developed in the population-based NOMASS protocol (NINDS ROl 29993). Demographic, socioeconomic, risk factor, and dietary data will be collected through direct interview of family members using standardized instruments. Blood will be collected for fasting glucose, electrolytes, creatinine, total serum homocysteine and related metabolites, and lipid profiles. All subjects will have high-resolution B-mode ultrasound imaging to measure carotid IMT and distensibility, brachial endothelial reactivity, and left ventricular mass. Storage of buffy coats and extraction of DNA will be done by the Columbia University Genome Center and genotyping will be done in year four-five by the Center for Inherited Disease Research. Statistical analysis will be conducted at Columbia University. Heritability will be evaluated using a pedigree-based maximum-likelihood method. Preliminary linkage analysis will be performed to identify quantitative trait loci linked to these phenotypes.